CN117136030A

CN117136030A - Perfusion measurement using low field NMR

Info

Publication number: CN117136030A
Application number: CN202180096806.3A
Authority: CN
Inventors: 谢尔盖·奥布鲁切科夫; 艾丽斯·利特尔
Original assignee: Willumio Ltd
Current assignee: Willumio Ltd
Priority date: 2021-04-09
Filing date: 2021-12-16
Publication date: 2023-11-28
Also published as: WO2022214865A1; AU2021440013A1; EP4319628A1

Abstract

The present disclosure relates to the field of medical applications. And in particular to a measurement system and method for non-invasively detecting and monitoring physiological parameters, particularly perfusion, to provide clinically relevant information by Nuclear Magnetic Resonance (NMR) observation.

Description

Perfusion measurement using low field NMR

Technical Field

The present disclosure is generally, but not necessarily, in the field of medical applications and relates to devices and methods for non-invasively detecting and monitoring physiological parameters, particularly perfusion, to provide clinically relevant information by Nuclear Magnetic Resonance (NMR) observation.

Background

All living tissues of the human body require a continuous supply of fresh blood, oxygen and other nutrients critical to the function of each cell. To meet this continuing need, blood is pumped throughout the circulatory system and delivered to the capillary bed where efficient transfer can take place, a process called priming. In terms of volume, perfusion generally refers to the rate at which blood passes through the capillary bed and is related to the metabolic demand of local tissue.

The use of perfusion measurements in stroke diagnosis, assessment and monitoring has attracted considerable interest, the main hazard mechanism of stroke being the sustained limitation of local tissue perfusion-leading to cell death and the formation of infarcts. Such direct association may allow early identification of the affected area to occur prior to many other lesion markers and provide a low level of clinical insight into the current status of any potential infarct and inform of the appropriate response.

However, perfusion measurements are difficult to implement on a large scale due to the limitations of the perfusion sensitive techniques currently available, which may be due to high system costs limiting usability (MRI, PET), the use of ionizing radiation to the detriment of additional imaging (CT, PET) or the difficulty of penetrating the skull (ultrasound, laser doppler). These problems severely limit diagnostic and evaluation applications and almost completely preclude any form of semicontinuous, informed perfusion, acute phase recovery monitoring.

Many forms of Magnetic Resonance Imaging (MRI) perfusion imaging have been developed, improved and implemented in routine clinical applications. Their disadvantage is that expensive non-portable machines are required to perform the test. Furthermore, a tracer-based method using contrast agents, such as gadolinium-based contrast agents, is often required. (GBCA). These are undesirable for a number of reasons and some have suggested limiting their use. MRI techniques are known for perfusion imaging without the use of contrast agents. Arterial Spin Labeling (ASL) is a more widespread alternative to MRI perfusion measurement techniques based on exogenous tracers. These sequences utilize blood in the body as a temporary tracer without the need for injection of contrast agents. The main limitation of these techniques comes from the relatively low signal-to-noise ratio-high resolution images are often not feasible due to the lengthy acquisition time. Thus, these techniques appear to be unusable in low field systems.

Disclosure of Invention

It is an object of the present disclosure to provide an improved non-invasive perfusion measurement and/or monitoring device. Additionally or alternatively, it is an object of the present disclosure to provide a useful alternative to known methods, devices or systems.

In one aspect, the present disclosure provides a perfusion measurement system comprising:

a magnetic structure configured to receive a body part of a subject under examination and to generate a non-uniform static magnetic field within a test tissue of the subject under examination;

means for applying a first reverse recovery Radio Frequency (RF) pulse sequence and a second reverse recovery RF pulse sequence to the test tissue, wherein the reverse pulses of the first pulse sequence or the second pulse sequence have a different bandwidth than the corresponding reverse pulses of the other of the first pulse sequence or the second pulse sequence; and

an acquisition device for acquiring magnetic resonance signal data from the pulse sequence and processing the magnetic resonance signal data to provide perfusion data.

In one embodiment, the signal data from the pulse sequences are compared to provide an estimate or measurement of at least one of flow, velocity, or perfusion.

In one embodiment, each reverse recovery RF pulse sequence includes a reverse pulse, an excitation pulse, and a Carr-Purcell-Meiboom-Gill (CPMG) acquisition.

In one embodiment, the bandwidths of the reverse recovery RF pulse trains other than the reverse pulses are substantially the same.

In one embodiment, the inversion pulses of the first inversion recovery RF pulse train have a narrower bandwidth than the inversion pulses of the second inversion recovery RF pulse train.

In one embodiment, the first reverse recovery RF pulse sequence affects the acquisition volume within the test tissue.

In one embodiment, the inversion pulse of the second inversion recovery RF pulse sequence is configured to invert NMR spins in a control volume (control volume).

In one embodiment, the control volume exceeds the collection volume.

In one embodiment, the control volume is configured to provide inversion buffer (inversion buffer).

In one embodiment, the inversion pulse comprises or is configured as a composite pulse.

In one embodiment, the narrow bandwidth pulse train comprises a flow-sensitive (flow-active) train, and the pulse train with the wider bandwidth inversion pulses comprises a flow-resistance (flow-resistance) pulse train.

In one embodiment, the narrow bandwidth pulse sequence comprises a marker sequence and the pulse sequence with the wider bandwidth inverted pulse comprises a control sequence.

In one embodiment, the control sequence collects data of a control volume of the test tissue, and the marker sequence collects data of a collection volume, wherein the collection volume is within the control volume. Preferably, the control volume is sufficiently greater than the acquisition volume to provide inversion buffering.

In one embodiment, the CPMG acquisitions are summed to improve the signal to noise ratio.

In one embodiment, a baseline pulse train is applied to the test tissue prior to the first and second pulse trains to provide a sampling unit.

In one embodiment, the sampling unit is repetitive. Preferably, the results from the sampling units are averaged.

In one embodiment, the inversion pulse is provided using a composite pulse. Preferably, the composite pulse comprises two 90 degree pulses separated by a 180 degree pulse, and most preferably comprises a 90 degree x pulse, a 180 degree y pulse, and another 90 degree x pulse.

In one embodiment, the bandwidth of the pulse train is configured according to the traffic. Preferably, the inversion recovery time is selected according to the type of perfusion fluid.

In one embodiment, the test tissue comprises a body part or organ. Preferably, the body part or organ comprises the brain.

In one embodiment, perfusion measurements are provided in real-time.

In one embodiment, the magnetic field strength is less than 1 tesla. More preferably, the magnetic field strength is less than 0.5T, more preferably about 0.25T.

In one embodiment, the system is single-sided.

In one embodiment, the perfusion measurement system is portable. Preferably, the portable system weighs less than about 30kg. More preferably, the portable system weighs less than about 25kg.

In another aspect, the present disclosure provides a method for measuring perfusion, the method comprising:

applying a non-uniform static magnetic field to test tissue of a subject under examination;

applying a first reverse recovery RF pulse sequence and a second reverse recovery RF pulse sequence to the test tissue, wherein the reverse pulse of the first or second pulse sequence has a narrower bandwidth than the corresponding reverse pulse of the other of the first or second pulse sequence; and

magnetic resonance signal data is acquired from the pulse sequence and processed to provide perfusion data.

In one embodiment, the method further comprises comparing signal data from the pulse train to provide an estimate or measurement of at least one of flow, velocity or perfusion.

In one embodiment, each reverse recovery RF pulse sequence includes a reverse pulse, an excitation pulse, and a CPMG acquisition.

In one embodiment, the method further comprises configuring the bandwidths of the reverse recovery RF pulse trains other than the reverse pulses to be substantially the same.

In one embodiment, the method further comprises configuring the inversion pulses of the first inversion recovery RF pulse train to have a narrower bandwidth than the inversion pulses of the second inversion recovery RF pulse train.

In one embodiment, the method further comprises configuring a bandwidth of the first reverse recovery RF pulse sequence to affect an acquisition volume within the test tissue.

In one embodiment, the method further comprises configuring a bandwidth of the inversion pulses of the second inversion recovery RF pulse sequence to invert NMR spins in the control volume.

In one embodiment, the control volume exceeds the collection volume.

In one embodiment, the method further comprises configuring the control volume to provide inversion buffering.

In one embodiment, the method further comprises configuring the inversion pulse as a composite pulse.

In one embodiment, the composite pulse comprises two 90 degree pulses separated by a 180 degree pulse. Preferably, the CPMG acquisitions are summed to improve the signal to noise ratio.

In one embodiment, the composite pulse includes a 90 degree x pulse, a 180 degree y pulse, and another 90 degree x pulse.

In one embodiment, the method further comprises the step of determining whether reduced perfusion occurs compared to a clinically acceptable level.

In one embodiment, the subject is a human.

In another aspect, the present disclosure provides a method of diagnosing reduced blood flow or perfusion in a subject below a clinically acceptable level using the measurement system defined in the above embodiments, the method comprising: applying a non-uniform static magnetic field to a test site of a subject;

applying a first reverse recovery RF pulse sequence and a second reverse recovery RF pulse sequence to the test site, wherein the reverse pulse of the first or second pulse sequence has a different bandwidth than the corresponding reverse pulse of the other of the first or second pulse sequence;

Acquiring magnetic resonance signal data from the pulse sequence and processing the magnetic resonance signal data to provide perfusion data; and

it is determined whether the perfusion data at the test site of the subject falls below a clinically acceptable level.

In one embodiment, the subject is a human.

In one embodiment, the test tissue to be examined is the brain, breast, kidney, liver or skin of the subject. In another embodiment, the test site to be examined is the brain, breast, kidney, liver or skin of the subject.

The present disclosure is described below with reference to specific embodiments. However, other embodiments than the above described are equally possible within the scope of the disclosure. It is within the scope of the present disclosure that method steps other than those described may be provided, with the method being performed by hardware or software. The different features and steps of the disclosure may be combined in other combinations than those described.

Other aspects of the disclosure will become apparent from the following disclosure.

In this specification, when reference is made to external sources of information (including patent specifications and other documents), this is generally to provide a context for the description of the features described. Unless otherwise indicated, reference to such sources of information in any jurisdiction should not be construed as an admission that such sources of information are prior art or form part of the common general knowledge in the art.

As used herein, the term "clinically acceptable" refers to standards commonly accepted or understood by medical practice, medical practitioners or clinicians.

As used herein, the term "about" in connection with a reference numeral designation refers to the reference numeral designation plus or minus up to 10% of the reference numeral designation. For example, the language "about 30" kg covers a range of 33kg to 27 kg.

As used herein, the term "and/or" means "and" or both. As used herein, "a" or "an" preceding noun refers to the plural and/or singular form of the noun. The term "comprising" as used in this specification means "including" or "at least partially consisting of. When interpreting statements in this specification which include that term, the features recited in each statement that begin with that term need to be present, but other features can also be present. Related terms such as "comprise" and "comprising" should be interpreted in the same manner. The entire disclosures of all applications, patents and publications cited above and below, if any, are incorporated herein by reference.

Drawings

One or more embodiments or implementations of the present disclosure will be described below with reference to the accompanying drawings, in which:

FIG. 1 is a schematic diagram of an embodiment of a non-invasive monitoring or detection system;

fig. 2-6 are isometric views of a stage of one embodiment of a magnetic structure.

Fig. 7A and 7B are magnetic field diagrams of the magnetic structures of fig. 2-6.

Fig. 8 is an isometric projection of a Radio Frequency (RF) coil support and conditioning apparatus for use with the magnetic structure of the preceding figures.

Fig. 9 is a graph showing the recovery of the bulk longitudinal magnetization (bulk longitudinal magnetisation) Mz after the inversion pulse, illustrating how a carefully timed pulse can cancel the signal due to the affected rotation, thereby potentially isolating the target signal.

Fig. 10 shows transverse and longitudinal magnetization diagrams (X-Z plane), resulting from long pulse lengths and short pulse lengths.

Fig. 11 shows a general pulse sequence profile (outline) of the ASL system disclosed herein. The implementation of the initial inversion may vary according to the disclosed requirements.

FIG. 12 shows the non-uniform flow sensitive interactive inversion recovery (IFAIR) sequence response over the entire range of the pump. T measured in undoped distilled water ₁ =2135 plus or minus 5ms. (A) 100us tag inversion. (B) 200us tag inversion. Sharing experimental parameters: 6000ms TR, 1600ms TI, 1200usTE,16 scans, 30us Control inversion.

Fig. 13 shows a performance comparison of an ifir sequence between using standard single pulse inversion and composite pulse inversion. T measured in undoped distilled water ₁ =2135 plus or minus 5ms. (A) full flow range. Highlighting of the flow decrease trend. Experimental parameters: 6000ms TR, 1600ms TI, 1200us TE, 16 scans, 100us label inversions, 30us control inversions.

FIG. 14 shows simulation data comparing Mz,0 generated by a standard single pulse inversion with a 90x-180y-90x composite pulse inversion.

FIG. 15 shows the use of different samples T ₁ Comparison between the observed trends. Specific parameters are as follows: (A) TR:600 ms, TI:1300ms (B) TR:3000ms, TI:400ms (C) TR:1500ms, TI:200ms. General parameters: 1200us TE,32 scans, 200us label reversal, 30us control reversal.

Fig. 16 shows a comparison of flow trends for a tube oriented along the long axis (y-axis) of the uniform region (sweet spot) and oriented perpendicular to the long axis (x-axis). The y-axis configuration is default and is used for other experiments. T measured in CuSO4 doped water ₁ =353 plus or minus 5ms. Experimental parameters: 1500ms TR, 200msTI,2500us TE,32 scans, 30us control inversions.

Fig. 17 shows an inversion Time (TI) scan at a fixed flow rate. T measured in CuSO4 doped water ₁ =320 plus or minus 8ms. Experimental parameters: 2000ms TR, 2500us TE, 8 scans, 500us label inversions, 30us control inversions.

Fig. 18 shows a graph of signal differences of inversion Time (TI) scan at a fixed flow rate. From the same experimental setup as in fig. 17.

Fig. 19 shows the simulated behavior of the signal difference in the TI range at a fixed flow rate. Vertical markers represent simulated T ₁ Values.

Fig. 20 is a comparison between ifir results before and after application of inverse quality scaling. T was measured in CuSO4 doped water using foam tissue phantom ₁ =320 plus or minus 10ms. Experimental parameters: 2000ms TR, 300ms TI, 2500us TE, 64 scans, 500us label inversion, 40us control inversion.

Fig. 21 shows a plot of spin substitution fraction versus flow rate. T was measured in CuSO4 doped water using foam tissue phantom ₁ =320 plus or minus 5ms. (A) 250us tag inversion. (B) 500us tag inversion. Sharing experimental parameters: 2000msTR, 2500us TE, 32 scans, 40us control inversions.

FIG. 22 shows a plot of spin substitution fraction per second versus flow rate. (A) 250us tag inversion. (B) 500us tag inversion. From the same experimental and simulated setup as in fig. 21.

Fig. 23 shows a plot of estimated flow rate versus directly measured flow rate derived from NMR signals. (A) 250us tag inversion. (B) 500us tag inversion. From the same experimental setup as in fig. 21.

FIG. 24 shows uncompensated T ₁ (tag sequence) and traffic Compensation T ₁ (control sequence) estimation. (A) 250us tag inversion. (B) 500us tag inversion. From the same experimental setup as in fig. 21.

Fig. 25 shows the signature and control sequence signals over time, measured in vivo, and scaled to eliminate effects due to inversion quality. Experimental parameters: 2000ms TR, 300ms TI, 2500us TE, 8 scans, 500us label inversion, 40us control inversion.

Fig. 26 shows an estimate of the spin substitution fraction per second calculated from scaled in vivo experimental data. From the same experiment as in fig. 25.

FIG. 27 shows flow compensated and uncompensated volume T calculated from scaled in vivo experimental data ₁ And (5) estimating a value. From the same experiment as in fig. 25.

Fig. 28 shows the magnetic structure of fig. 2 to 6 and 8.

Fig. 29A and 29B show the results obtained for the magnetic structure of fig. 28 for the permanent set (permanent damage model) (fig. 29A) and the transient set (transient stroke model) (fig. 29B).

Detailed Description

The devices or apparatus described in this specification may be used in a variety of applications, some of which may be medical independent. However, to describe the disclosed apparatus and/or methods, the present disclosure will be applied, by way of example, in a device capable of detecting or monitoring flow, particularly blood flow in tissue in a selected region of the human or animal body. These regions may be, but are not limited to, limbs and/or organs (e.g., brain).

The brain is one of the vital organs of noninvasive technology. Brain tissue health, i.e. parameters of brain oxygenation, blood perfusion and diffusion (used as markers of cell damage), often requires information. An apparatus possible according to the present disclosure may operate based on an application of time domain relaxation measurements using Nuclear Magnetic Resonance (NMR) techniques. A significant advantage is that one or more of the magnetic structures disclosed herein use low field strengths (0.25T), resulting in a magnetic field strength of between about 2 and 5cm ³ Small uniform areas (best points) in between.

The configuration of the magnetic field is not designed for Magnetic Resonance Imaging (MRI). MRI magnets are very high field strength magnets that are specifically configured for imaging applications. The magnetic structure as described herein is unique in reduced field strength, which sacrifices imaging capability to provide near real-time information measurement or monitoring information.

The present disclosure allows for the construction of devices that have a significantly reduced form factor and provide portability in systems weighing about 25-30 kg. By reducing the size of the magnetic structure, the cost of manufacturing such devices can be significantly reduced while still providing clinically valuable information.

Referring to fig. 1, an example of a system, generally designated 1, according to one embodiment of the present disclosure is schematically illustrated. The system 1 includes a magnetic structure, generally designated 10, a controller 20, and a user interface 30. The magnetic structure 10 includes a coil 12, and the arrangement and operation of the magnetic structure 10 and the coil 12 will now be described in further detail below.

The controller 20 may include more than one component or module. For example, the power source for powering the coil 12 may be provided separately or may be provided as part of the controller 20 or the coil 12. The controller typically comprises a computer module or microprocessor, or Field Programmable Gate Array (FPGA) or similar device programmed or configured with software to execute control instructions contained in the software to perform the necessary control functions as described further below.

In some embodiments, the processor for performing the control functions may take the form of a general purpose computer, such as a laptop or tablet computer. Such a device may also provide a user interface 30, which may for example comprise a touch screen.

As can be seen in fig. 1, the magnetic structure 10 is designed to provide a constant magnetic field and is of sufficient size to receive or receive a head (such as a human head), which is shown at outline marker 14. In the figure, the head 14 is seen from the top of the head, in other words, the patient is prone along a longitudinal axis directed towards the page (according to the y-axis of the drawing).

Portable magnet systems (portable Magnetic Resonance (MR) systems) according to the present disclosure, such as the magnetic structure 10, have been designed and constructed to allow NMR to be used to detect and monitor tissue parameters in organs such as the human brain. In particular, a system according to the present disclosure may measure T2 changes due to blood oxygenation in the brain. This requires strong magnet optimized B ₀ The field strength, with low form factor, high homogeneity (low magnetic field gradient) and a sufficiently deep homogeneous region (optimal site), can reach most areas of the brain. The term "homogeneous region", "optimum site" or "homogeneous region (optimum site)" in this sense refers to a region of substantially uniform field strength throughout a tissue test volume sufficient to detect or monitor one or more parameters by NMR. Referring to fig. 1, the shape and direction of the magnetic field generated by the magnetic structure 10 is shown by arrow 16. The field provides a uniform region (optimal site) 18 within the patient's head 14. The size of the uniform region (optimum spot) can be varied by adjusting the magnetic assembly, as will be described further below. In one embodiment, the site is about 10mm wide, about 10mm high and about 50mm along the y-axis. In other embodiments, the dimensions of each dimension vary from about 5mm to 50mm.

The magnetic structure 10 is designed to fit or accommodate the head of an average person while keeping the face accessible. The magnetic structure 10 and the system as a whole are also designed to be compact and light enough to be easily moved. Those skilled in the art will appreciate that the magnetic structure 10, which is sufficient to accommodate a person's head, will also be capable of accommodating other body parts or organs, such as many joints and limb parts, and possibly be placed around a portion of the torso to enable non-invasive detection or monitoring of tissue of various different organs.

The components of the magnetic structure 10 will now be described. In the illustrated embodiment, the design of the magnetic structure is constrained by the maximum and strongest magnets available from the manufacturer. The main magnet assembly uses eight 90x90x50mm (wide x deep x high) neodymium magnets. The optimization of magnet placement by simulation shows that the design can produce a uniform area (best point) of 0.25T at about 30mm above the magnet surface. This site is about 10mm wide by about 10mm high and about 50mm along the y-axis. Those skilled in the art will appreciate that the magnetic structure 10 itself may be tuned to change the size of the uniform region (optimum site), as described further below. Moreover, the magnetic structure 10 has been configured to allow for a change in the position of the body part it receives relative to the magnetic structure in order to adjust the position of the uniform region (optimal site) in the observed tissue. This also includes being able to change the position of the coil 12, as described further below.

In the illustrated embodiment, the performance of the magnetic structure 10 is simulated at COMSOL Multiphysics (Burlington MA, USA) using a finite element method to solve the laplace equation for magnetic scale. This allows an accurate modeling of the effect of the high permeability yoke. The magnet design can be refined by scanning different values of the design parameters in COMSOL and optimizing the generated field for uniform region (best-spot) location, intensity, and field uniformity.

Referring to fig. 2, other aspects of the magnetic structure 10 are shown, including a non-magnetic (e.g., aluminum) center support 110 mounted on a yoke base 100 made of magnetic steel (e.g., 1010, 1016 or the like) that has been milled to specification. The side support 120 is attached to the yoke 100, followed by a rail 130, which ensures the base magnet is correctly positioned. Top bracket 150 secures or supports the distal or upper ends of supports 110 and 120.

As shown in fig. 3, the magnet base 160 provides additional rigidity to the base to prevent the magnetic structure, including the yoke base 100, from bending under the influence of the magnet.

Other aspects of the magnetic structure 10 are shown in fig. 4-6. These figures also illustrate coil assembly 12, which will be described further below with reference to fig. 8.

The base magnet 170 is received or secured in place between the supports 110 and 120 near the bottom region of the support, i.e., the base magnet 170 is disposed proximate or adjacent to the base 100. The upper magnets or wing magnets 190 are located in the upper region of the supports 110, 120 such that they are located away from the base 100.

The completed magnetic structure has two sidewalls a and B (as shown in fig. 1 and 1A) formed by the supports 110 and 120, or by the magnets 170, 190, or by a combination of the supports 110, 120 and the magnets 170, 190. In one embodiment, walls a and B may be parallel to each other, or substantially parallel to each other, or angled to each other. In other embodiments, such as the illustrated embodiment, walls a and B are angularly spaced from each other forming a channel 192 (shown in phantom in fig. 1A), whereby the mouth 194 of the channel (at the top of the magnetic structure) is wider than the base 196 of the channel.

In some embodiments, such as the illustrated embodiment, the longitudinal (i.e., along the y-axis of fig. 1) extent or dimension of the base magnet and the wing magnet is greater than the transverse (i.e., along the x-axis of fig. 1) extent or dimension thereof. In some embodiments, the base of one or both of walls a and B has a greater lateral dimension at the base than at the top, i.e., one or both walls are thicker at the base than at the top, such that they are thicker at the base 196 closer to the channel 192 than at the mouth 194 of the channel 192. In some embodiments, such as the illustrated embodiment, walls a and B include a plurality (two or more) of individual base magnets 170 and wing magnets 190 along the longitudinal (i.e., y-axis in fig. 1) direction.

Fig. 7A shows the COMSOL simulated magnetic field of the device of the previous figures. Fig. 7B shows the magnetic field generated by the magnetic structure according to the embodiment shown in fig. 2-6, measured using an internally constructed 3-axis field mapping system (mapper system).

The aluminum frames 120, 130 hold the magnets 170, 190 in place with the steel yoke 100 and are configured to hold the magnets in place and resist 700N attractive forces affecting the two wing magnets 190. The frame structure spaces the magnets apart and provides a channel 192 that includes an open space into which a body part may be received for non-invasive analysis, detection or monitoring. In some embodiments, the supports 110 and 120 are slidably and/or pivotally mounted relative to the base 100 such that the channel may widen (or narrow) to provide more (or less) access space by moving the wing magnets 190. If the magnets are moved farther apart, more space is provided to accommodate a larger head or other body part, but at the cost of a slightly reduced field strength and location of the uniform region (optimal site). Moving the wing magnets 190 closer together will increase the field strength and increase the location of the homogeneous region (best point). Thus, the magnets can be adjusted to alter the uniform region (optimal site) as needed to locate and/or alter the size and field strength of the uniform region (optimal site) relative to the tissue.

As described above, NMR systems use a constant magnetic field (provided by magnets 170, 190 in this embodiment), but also rely on a weaker oscillating field. Referring to fig. 8, the oscillating field is provided by a Radio Frequency (RF) coil, such as coil 200 of coil assembly 12, which is used to generate B ₁ The magnetic field excites the spins and detects the transverse magnetization of the spins. The RF coil 200 and associated electronics disclosed herein are similar to other NMR and MRI instruments.

The RF coil assembly 12 used in some embodiments disclosed herein consists of an RF coil 200 that is actually an RLC circuit that resonates at the larmor frequency of protons at the magnetic field of the uniform magnetic region. The circuit was tuned and matched using a variable capacitor to achieve a Voltage Standing Wave Ratio (VSWR) value of 1:1.43 (return loss 15 dB). The pick-up coil 210 is a 3-turn oval solenoid made of a tinned copper wire having a diameter of 1mm and is wound on a 3D printing member adapted to accommodate the shape of the head. The RF coil 200 is carried on telescoping or length adjustable struts 210 and 212 and can be moved up and down relative to a base platform 214 as indicated by arrow 216 to accommodate objects of different sizes. The RF coil assembly 12 may also be moved to change the spatial position of the acquired signal, such as the penetration depth.

The shim magnet assembly 218, which in some embodiments is disposed on the bottom below the base 100, moves up and down and side-to-side to adjust the magnetic field B _o Is a uniform property of (a).

In some embodiments, the shim magnet assembly 218 is located in two trays below the RF coils and is attached to a movable platform 214 that can translate the shim magnet assembly 218 up, down, and side-to-side in order to improve B ₀ Uniformity of the magnetic field. The improvement in uniformity can be measured in a number of ways, but is typically measured using either 1) a 3-axis field mapper with a hall effect probe or 2) NMR techniques using, for example, a large sample of long doped water and observing an effective T using a karl-pezier-Mei Bum-gil (CPMG) experiment ₂ 。

The shape of the magnet assembly is configured to partially surround the RF coil, thereby limiting the electromagnetic flux to which the RF coil 200 is exposed, thus partially shielding the coil 200 to reduce electromagnetic noise. This makes it very robust to any external interference. In some embodiments, further noise reduction uses a conductive RF shielding textile material to cover the subject/patient to further reduce electromagnetic interference. In some embodiments, an additional denoising strategy using active noise cancellation of the pickup RF coil may be used to further improve the signal-to-noise ratio.

Suitable NMR console electronics (such as those provided by the new zealand wheatstone resin company limited) include an RF transceiver, serial and parallel I/O lines for running flip-flops for temperature monitoring and control, an RF power amplifier (such as provided by TOMCO of Stepney SA, australia), a software code sequence for fast signal calibration.

It has been found that low field systems can provide perfusion measurements. This is achieved by carefully controlling the area where sampling and detection occurs to improve the effective signal quality. In a sense, the system disclosed herein "images" only a single pixel.

The pulse sequences used have some similarities to the Arterial Spin Labeling (ASL) technique used by conventional MRI systems. This known technique, known as flow-sensitive interactive inversion recovery (FAIR), is a variant of the known Pulse Arterial Spin Labeling (PASL) sequence. The FAIR procedure compares the signal responses of two slightly different reverse recovery pulse sequences. Each sequence consisted of 180 degree inversion pulses followed by image acquisition after a short delay. The first sequence uses slice selection gradients to limit the initial 180 degree pulse to only reverse the spins in the imaging plane. The second sequence does not apply a gradient but allows a 180 degree pulse to affect the entire sensitive volume. As shown in fig. 9, by carefully timing the delay between inversion and image acquisition, the signal produced by the inverted spins can be largely suppressed. The 180 degree pulses of the first sequence do not interact with blood that is about to flow into the imaging plane. Thus, the observed signal difference between these two sequences will be related to the rate of new blood entering the tissue.

While there are many competing ASL protocols that have been published and validated, these protocols are designed for higher field medical MRI systems, with large sensitive volumes and comprehensive gradient coils. The 9MHz non-uniform field NMR system disclosed herein does not possess any of these features, and therefore any protocol implementation must be re-conceived to function without these features.

Non-uniform field NMR systems are typically not equipped with gradient coils because multi-pixel imaging is not within the scope of such systems. Thus, standard spatially dependent measurement protocols are typically not directly convertible from higher field systems.

However, it has been found that naturally occurring gradients of non-uniform fields can be exploited.

As described above, the 9MHz prototype magnet was a homogeneous zone (best-spot) based system. In the center of the homogeneous region (optimum position), the magnetic field is relatively uniform, B ₀ The gradient increases gradually with increasing distance from the center of the homogeneous zone (optimum locus). Since the uniform region (optimum locus) is not suddenly cut off between uniform and non-uniform, the effective size of the uniform region (optimum locus) depends on the frequency bandwidth of the applied excitation pulse. This allows for the application of a wide bandwidth pulse or respectively A narrow bandwidth pulse excites a large volume or a small volume. The shape of the excitation volume is shown in fig. 10, showing the dramatic effect of the X-shape and pulse bandwidth.

The bandwidth of the excitation or inversion may be operated by changing the shape, duration or form of the RF pulse. For example, the Shinnar-Le Roux (SLR) pulse is shaped to reduce bandwidth and allow finer frequency selection.

For standard hard pulses, the bandwidth is inversely related to the duration. By compensating for the RF amplitude, the two hard pulses have equivalent flip angles, but the frequency distribution is significantly different. In addition to a single RF pulse, the bandwidth of the inversion can be increased by using some composite pulses.

The composite pulse is a series of RF pulses that together achieve a particular implementation

Effects. The composite pulse used in at least one example or implementation consists of two 90 degree x-pulses and a 180 degree y-pulse in between, together creating an inversion. Within this composite pulse, the rotation caused by the 180 degree y pulse results in cancellation of errors in the flip angles of the two 90 degree x pulses associated with field non-uniformities. This corrects off-resonance spin, resulting in a wider effective bandwidth and more uniform inversion than is achieved by equivalent single pulse inversion.

Pulse sequences have been designed to achieve a similar FAIR

The effect of the protocol, but in a non-uniform field. For convenience of description in this disclosure, this new measurement procedure is referred to as non-uniform flow sensitive interactive inversion recovery (ifiir). The ifir sequence adopts an inversion recovery mode: the pulse is reversed, followed by the excitation pulse, and then the karl-pezier-Mei Bum-gill (CPMG) harvest.

Two slightly different reverse recovery RF pulse sequences are defined. The first sequence may be referred to as a tag sequence. The sequence uses a uniform narrow bandwidth (or at least a bandwidth narrower than the bandwidth of the inverted pulse of the second pulse sequence) for all pulses, including the initial inversion. The second reverse recovery RF pulse sequence may be referred to as a control sequence. The sequence is substantially the same as the first sequence except that the bandwidth of the initial inversion is modified to be wider than the inversion pulses of the first sequence, preferably by using shorter and stronger RF pulses. The exact implementation of the inversion pulses in the two sequences is different, since the bandwidth modification options disclosed above can be explored. The CPMG echoes are summed to improve the signal to noise ratio. Fig. 11 shows the structure and relative timing of the pulse train.

The ifiir protocol detects flow rate by analyzing the signal response difference between the first pulse sequence and the second pulse sequence (i.e., by comparing the signals).

As described above, the bandwidth of the inversion pulses of the second pulse sequence is wider than the bandwidth of the inversion pulses of the first sequence. Thus, the inversion pulse inverts the spins in a large volume of target tissue, which may be referred to as a control volume. The control volume includes the acquisition volume and extends beyond the acquisition volume by a margin to provide inversion buffering. Since the excitation and acquisition segments of the second sequence are configured to be of narrower bandwidth than the first sequence, the acquisition is still sensitive to the control volume. Notably, because the spins in the control volume have been reversed, the spins of any fresh fluid that entered the control volume during the execution sequence are not detected. Thus, the second sequence can be considered a flow insensitive control sequence.

The first pulse sequence affects the acquisition volume within the target tissue by having a narrower bandwidth. The inversion pulse is configured to suppress spins in the acquisition volume such that substantially only non-inverted spins from fresh fluid (e.g., blood) are detected by the acquisition when the pulse sequence is performed. Thus, the first sequence may also be considered a stream sensitive sequence, or a tag sequence.

The order of the first pulse sequence and the second pulse sequence may be changed, i.e. the second sequence may be performed before the first sequence. The pulse sequences are preferably performed one after the other.

Thus, a comparison of the signals acquired from the first pulse sequence and the second pulse sequence should reveal a difference indicating that fresh fluid has entered the acquisition volume. However, due to field inhomogeneities, it may be difficult to detect a comparison between the affected volumes of contrast inversion and label inversion, requiring careful consideration of applicable flow scenarios.

In order to maintain the clarity of this disclosure, it is necessary to define several terms relating to the various features of the pulse sequences described above. These terms are given in table 1 below, which will aid in the subsequent explanation of experimental development and analysis.

/>

TABLE 1

In practice, these two complementary sequences (i.e., the control sequence and the marker sequence) are performed consecutively, forming part of the experimental unit. The unit is repeated multiple times and averaged to improve the signal to noise ratio. The complete repeat unit also includes non-inversion acquisition that provides a baseline, fully recovered signal S _∞ . Since the marker sequence and the control sequence differ only upon initial inversion, S _∞ Can be used as a baseline signal for both. To achieve the improved processing methods disclosed further below, the marker sequence and the control sequence may also be repeated, each sequence having a very short inversion Time (TI), providing a measure of the system after inversion but before significant recovery can occur.

Table 2 below summarizes the different signals collected in each experimental unit and the corresponding symbols.

TABLE 2

Interleaving these different measurements ensures that any systematic changes that occur during the experiment will have approximately the same effect on each output signal, rather than unpredictably distorting the results.

Baseline readings allow for calculation of normalized signals(see equation 1 below) to improve consistency between experiments. This normalization corrects for any shift or drift in signal strength due to sample characteristics and position or magnet temperature, allowing better quantification of results.

Equation 1

To assist the addressee (address), a general mathematical description of the signal response of the ifiar sequence is listed below.

Two related contributions are T ₁ Restoring and effectively replacing spins within the acquisition volume due to flow. All other effects can either be controlled for consistency such as sample position or be insignificant in comparison.

Equation 2 is the most general equation describing the effect of these factors on the normalized observed signal

Equation 2

Where eta (v, t) is a function of the fraction of the signal contribution describing the spin, which is replaced in the inversion time t due to the average flow velocity v, Is the longitudinal magnetization before 90 degrees excitation at time t, normalized by the equilibrium magnetization. Eta and->The form of both functions depends on the parameters of the pulse sequence and the properties of the sample. Various possibilities can be proposed, modeled and verifiedAnd (5) solving.

Equation 2 can be applied to M _z,0 Special cases of calibration measurement. In this case, the flow dependence can be ignored, since a short inversion time means that any effect will be insignificant. Taking the zero time limit and substituting T of the Buloch equation ₁ Relaxation solution yields equation 3:

equation 3

Wherein the method comprises the steps of

This can be applied to both the marker and the control calibration and shows that the resulting signal can be used as a direct measure of the initial magnetization due to the corresponding inversion pulse.Can be understood as reversing mass; />Corresponding to a complete inversion in which each spin in the sample is exactly 180 degrees rotated from equilibrium. Based on the system calibration and the pulse parameters,may vary between measurements.

Motif-based experiments have demonstrated the rationale for the above-described ifiar procedure. Experiments have shown that the ifiir sequence can generate a signal that is dependent on the flow circuit flow rate. These experiments were performed on a simple system, without doping water, without tissue phantom and periodic single pulse inversion to keep confounding factors to a minimum.

Fig. 12 shows a characteristic signal flow curve over the entire available flow range at two different mark inversion pulse lengths. Clearly, there is a flow rate threshold above which the buffering of the reverse spin from the control reverse is insufficient to prevent recovery due to flow. Comparison of the two graphs shows that the control signal remains unaffected for a greater range of flows when using a narrower bandwidth 200ms pulse. This follows the expected behavior, as smaller acquisition volumes allow for greater inversion spin buffering against inversion.

The most useful part of these signal flow curves is the generally linear region observed at lower flow rates. In this area, no substantial recovery of the control signal occurs, thus maintaining the underlying principles behind the procedure. The range of flow rates in which this occurs depends primarily on the bandwidth of the pulse length used and the inversion recovery Time (TI) of the observed flow. These parameters can be optimized to create sequences that are sensitive to the amplitude of the associated velocity.

These figures highlight the problem of using the original signal differences to estimate the unknown flow, i.e., there is a signal mismatch between the zero flow signature curve and the control curve. This offset is caused by the non-uniformity of the initial inversion pulse mass and can be a barrier to qualitative and quantitative flow measurement. While the target overall signal flow trend is relatively unaffected, the offset may mask the relative recovery due to flow unless the system can also take measurements at zero flow conditions. This problem is further addressed in more detail below, in which an improvement to the pulse sequence is shown and a flow independent compensation method is presented to address this problem.

The proposed modification to the basic inversion recovery sequence is to use a composite inversion pulse. As described above, these pulses use a short series of individual pulses, theoretically achieving a more uniform inversion over a wider bandwidth. FIG. 13A shows a comparison between a pair of flow scan experiments, one acquired using a standard single pulse inversion sequence and the other acquired using 90x-180y-90x complex inversion.

These figures illustrate several advantages of composite pulses. First, the control signal shows greater recovery resistance due to flow, maintaining an initial zero flow response over a greater range of flows. This indicates that the bandwidth of the inversion is improved. Second, when a composite pulse is used, the zero flow response of the marker signal shows a better match with the control signal. Again, this can be attributed to an improvement in pulse bandwidth which more uniformly covers the acquisition volume. Fig. 13B shows how these improvements lead to signal differences starting from zero without flow at lower flows and steadily increasing with the introduction of flows.

Fig. 14 shows simulated longitudinal magnetization produced by standard pulses and composite pulses of different durations. These results were generated by a bloch equation simulator with a pulse interval of 2500ms. These figures show how the composite pulse significantly improves the range of near-optimal inversion occurrences, thereby improving inversion uniformity. However, the magnitude of this effect appears to depend on the initial pulse bandwidth and is therefore more pronounced for shorter, wider bandwidth pulses.

The results of the above discussion are all in T with extension ₁ The values were collected in undoped distilled water. T of blood and tissue ₁ This is particularly true under in vivo conditions, in contrast to pure water. Since the ideal inversion Time (TI) of the sequence depends on the T of the sample ₁ Thus helping to demonstrate these effects on the signal flow profile. FIG. 15 shows T ₁ Flow dependence at values 2135ms, 615ms and 310 ms. For matching, TI parameters of 1300ms, 400ms and 200ms were used, respectively. These values are chosen because they have good signal return-to-zero characteristics, corresponding to T ₁ About 65% of the value.

The results show that modified T ₁ The main effect of (2) is to reduce the sensitivity to convection. This is because shorter TI results in a corresponding decrease in spin displacement after the inversion pulse, effectively replicating lower flow behavior. No significant difference was observed other than this change. To achieve the goals of in vivo experiments, all subsequent experiments were performed with doped water to approximately match 0.19T B ₀ Related organization T of 350ms in field ₁ 。

The uniform region (optimal site) of the magnet has one long axis and two substantially equal shorter axes. Since the flow-related changes in the signal response are due to the substitution ratio of spins in the uniform region (optimum site), the direction of the flow will change the result. Most other experiments were performed with the flowtube aligned with the longer y-axis.

Figure 16 shows that similar flow dependence occurs when the tube is oriented along the shorter x-axis, but with much higher sensitivity. The control signal did not exhibit any flow-dependent recovery, indicating that despite the steeper gradient, the control volume was still large enough compared to the acquisition volume to create a usable inversion buffer.

The TI in the experiment was chosen to be a value where the signal contribution of the un-substituted spin is approximately zero (as described above). However, the TI value selected interacts with the measured flow sensitivity because the spin-shift distance during the experiment is directly dependent on the time at which the flow is observed. Furthermore, T is measured by a standard IR pulse sequence ₁ Is affected by the flow of the sample. In cases where no zero-flow calibration, such as in vivo experiments, can be performed, this may confuse the true T ₁ Is a function of the estimate of (2).

One potential solution to this calibration problem is to perform the ifir sequence at a series of TI values so that well-behaved values can be directly identified. Fig. 17 and 18 show how the marker and control signals vary with TI at 0 mm/sec (stationary) flow, 1 mm/sec flow, and 2 mm/sec flow. These figures show that the difference between the two signal responses remains relatively consistent over a larger TI range. When TI is approximately equal to T of the sample ₁ The greatest difference occurs when. FIG. 19 shows that when TI and T ₁ How the signal difference continues to peak at the time of matching.

The results can be verified by comparison with a model that considers two competing effects of TI on signal response. The low flow uniform length model described above is applied.

A complete inversion pulse will precisely flip each affected spin 180 degrees. Due to B ₀ And B ₁ Field inhomogeneities, which are not possible in practice-the effect of achieving this effect is called "marking efficiency" or "quality" of the inversion pulse and corresponds to a normalized instantaneous longitudinal magnetization

While the use of composite pulse inversion helps mitigate the difference in inverted pulse quality between the marker pulse and the control pulse, the underlying problem remains. This is especially true when longer, narrower bandwidth marker pulses are used to increase sensitivity to low-speed streams, as the more limited bandwidth reduces the effectiveness of the composite pulse (see fig. 14).

A more robust solution is to use the minimum TI calibration measurement described above for direct measurementAnd adjusting the signature and the control signal to compensate. />Variation pair->The effect of the time evolution of (and hence the recorded IR signal) is T through the bloch equation ₁ And (3) determining a relaxation solution. The equation shows that after excitation or inversion, the normalized longitudinal magnetization is calculated fromTime is restored to t>>T ₁ And +1. Due to->Highly dependent on RF frequency, bandwidth and B ₀ The interaction between the precise non-uniformities of the fields can give varying shifts in the observed signal, limiting the accuracy of the comparison between experiments.

This can be done by defining a transformationTo solve, the transformation will

The results of each experiment are mapped onto a consistently defined domain, correcting

Offset, as shown in equation 4.

Equation 4

Due to control T ₁ The characteristics of the recovered exponential function,

definition of the definitionThe relevant information is preserved for the time constant that a simple linear resampling of the new range does not affect the recovery curve. This eliminates the effects from signal flaw inversion, isolates the information related to the flow and artificially replicates the behavior of an ideal homogeneous field system. Equation 5 gives the new scaling signal +.>The correlation of the flow velocity v and the inversion time t.

Equation 5

Applying this to experimental data, fig. 20 shows how the difference curve of the scaled signal always starts from zero at zero flow and steadily increases from there, regardless of the zero flow offset of the non-scaled signal. The signal processing strategy depends on the accuracy of the calibration signal, but the baseline measurement (S _∞ ) And short TI measurementAnd->) Provides reasonably good signal to noise ratio.

One step in quantifying the flow rate will depend on spin substitution (T) ₁ Relaxed label and control signal S ^t And S is ^c The conversion is to describe only the amount recovered due to spin substitution.

This can be done using control of each signal pair spin substitution and T ₁ The equation for the recovered dependencies is shown. In this case, it is assumed that the flow rate-inversion time product is small enough that the control inversion covers all spins contributing to the control signal. This assumption is applied to equation 5 to yield equations 6 and 7 for the tag signal and the control signal.

Equation 6

Equation 7

By rearranging and arrangingSubstituted into->Spin substitution fraction η

May be represented by only the measured signal quantity as shown in equation 8.

Equation 8

If and only if the T of spin is shifted ₁ The rightmost simplification is valid only when the spins contributed by all signals are approximately common. This is effective for flow motifs but is unlikely to be applicable to real tissue. This general expression works for arbitrarily shaped acquisition volumes, but depends on the assumption that the control inversion is equally valid for all incoming spins. As previously mentioned, this is quite accurate at low traffic and small mark inversion bandwidths, but is not entirely positive And (5) determining.

Fig. 21 shows how this measurement varies with flow at different TI values. Longer TI experiments resulted in higher η values due to the increased displacement distance during the experiment. Comparing the two graphs shows that a narrower bandwidth 500ms experiment results in a larger recorded fraction of spin substitution. This is due to the reduction in acquisition volume size combined with the linear flow of the loop.

Fig. 22 shows the same data, but scaled by the measured TI.

This gives an estimate of the spin substitution fraction per second of inversion recovery time (defined as ζ=η/TI) and allows a direct comparison between the 3 curves shown on each graph. The 250ms mark pulse plot shows excellent agreement between 3 TI values at all measured speeds; the 500ms mark pulse plot shows excellent uniformity for flow rates below 1mm/sec, but slightly offset above this point. This difference may be due to the shape of the acquisition volume at this very narrow bandwidth, the longer TI spin shift partially penetrates the thin region and reduces the response ratio.

Fig. 21 and 22 also show the simulation set. A relatively good match between the simulated data and the actual data indicates that the bow of the ζ -velocity curve is due to a similar combination of fast-replaced and slow-replaced regions.

In the context of a tunnel-based tissue phantom, the spin of all contribution signals

Can be assumed to be a flow and the flow is unidirectional, the spin substitution per second fraction ζ is related to the flow rate. As described above, the exact relationship depends on the shape of the acquisition volume, but can be estimated by applying a simple model. The approach taken is to consider the shape as having a single flow alignment length and excitation and inversion effectiveness that is uniform throughout the volume.

Under this approximation, by using equation 9

Model flow rates may be estimated.

Equation 9

Fig. 23 shows a comparison of the estimated flow rate with the directly measured flow rate. The resulting estimate does follow the general trend. The appropriate length parameter of the UL model is indeed proportional to the marking bandwidth and thus also to the size of the acquisition volume; FIG. 23A is a diagram of the use of l _Model Generated with =4.4 mm, fig. 23B is with l _Model Generated =2.2 mm. This shows that in this case, when the pulse bandwidth is doubled, the average flow alignment length nearly doubles.

In a non-uniform field, T ₁ The standard Inversion Recovery (IR) measurement of (c) is affected by the sample flow. This is due to the same mechanism of observation and measurement by the ifir sequence, the spins within the acquisition volume are replaced by spins that are not affected by the initial inversion. The tag sequence of the ifiir protocol shown in fig. 11 is similar to the standard IR sequence, but differs significantly in the implementation of the initial inversion. This shows that, if the tag sequence corresponds to a standard, flow-affected IR sequence, at low flow rates,

The control sequence can be used to measure the true, static T of the active flow sample ₁ . This can be clearly seen in equations 6 and 7 set forth above.

Once the above processing technique is applied, T of the resulting signal ₁ The recovery will follow a simple exponential recovery procedure with only one unknown parameter. This allows estimation of T from a single point in time ₁ . This means that standard, stream-affected T can be extracted from the IFAIR sequence output ₁ And flow compensated T ₁ Both without additional measurement time.

FIG. 24 shows the estimated T during the same experiment presented in the previous section ₁ How to vary with flow rate. The graph clearly shows that at zero flow, both estimates are equal to T ₁ Steady state IR sequence results of =320 plus or minus 5ms match well and, within the measurement range, contrast the derived T ₁ Is not affected by the flow rate.

In order to demonstrate feasibility in clinical settings, in vivo experiments have been performed.

One experiment aimed at observing changes in blood perfusion of the arm muscle tissue of a human by repeatedly performing the above-described ifir procedure during alternating rest and muscle contraction. Due to the changing metabolic demand, the blood perfusion of the muscle should increase during contraction and return to baseline during rest.

Muscle contraction is achieved by actively squeezing a length of compressible foam as consistently as possible within 3 minutes. To accelerate the acquisition process, the mass measurement S is reversed _t,0 And S is _c,0 Recorded only once at the beginning of the experiment and scaled to match the baseline signal S during the experiment _∞ Any minor variation of (2). The pulse sequence timing was tr=2000 ms, te=2500 us, control was inverted to 40us and mark was inverted to 500us. TR is long enough for full recovery to occur between each scan. At each time point, S is collected _∞ 、S _t,0 And S is _c,0 Each 8 scans, ti=300 ms. The acquisition time for each complete scan is less than 60 seconds.

The arm was placed on a plexiglass plate, placed over the RF coil with the uniformity zone (optimal site) located approximately 25cm proximal to the wrist joint for most of the superficial flexor carpi found along the anterior portion of the forearm. The main target muscles are the superficial flexor and palmaris longus, as they are involved in the grip function and are located at the superficial position of the forearm. The center of the uniformity zone (optimal site) was estimated to be about 5mm from the skin surface. The arm and magnet are covered with a piece of conductive fabric to reduce signal noise.

Fig. 25 shows the evolution of the marker signal and the control signal over time, as well as indicating the muscle activation period. When the arm muscle is activated by grasping, a very significant increase in the marker signal is observed, but the control signal is not affected. Once the muscles relax, the signature signal quickly returns to near baseline values. The intensity of the marker signal during muscle activation is not always consistent, but this reflects the difficulty of maintaining a consistent level of muscle activation over a period of several minutes, resulting in gentle changes in local perfusion.

Fig. 26 shows the spin substitution fraction η per second estimated from the same data. In this case, T of blood and tissue ₁ Should be different, thus need to be specific to T ₁ Estimation of bloodGauge (see equation 8). Blood T at 0.2T ₁ Plus or minus 22ms for 775; this value is used for calculation. Since a single constant value is used for the conversion, the shape of the ζ curve matches the signal difference Δs. Regardless, the figure clearly shows that at rest, the blood perfusion of the forearm muscle is relatively low, replacing about one third of the spin of the contribution signal in the 500ms acquisition volume per second. These results indicate that the rate of blood exchange in the muscle capillary bed increases the resting value up to 5 times and is able to maintain this rate over a period of several minutes. This is consistent with the literature on muscle perfusion, where a substantial increase in the percentage of muscle perfusion during exercise is a recognized behavior.

FIG. 27 shows T derived from two IFAIR sequences ₁ Values. As with the flow motif, the control sequence produces a stable value that is not affected by the hypothetical flow changes, while the standard IR adjacent marker sequence values drop at higher flow rates. Static T of the experiment ₁ The value is equal to the T based on independent IR performed in the same time period ₁ Measurement results are very consistent-T ₁ =330 plus or minus 20ms.

The second in vivo experiment used a portable Magnetic Resonance (MR) system, as shown in fig. 1 to 6 and 8 and described in detail above, to study pathophysiological changes that occur after ischemic stroke in a sheep stroke model. The protocol of this experiment was approved by the university of aldehyd animal ethics committee (Animal Ethics Committee of the University of Adelaide).

In this experiment, 11 merino sheep were divided into two groups. The first group 6 is known to undergo permanent occlusion surgery (hereinafter referred to as the "permanent group" or "permanent injury model"). The second group 5 received only temporary occlusion surgery (hereinafter referred to as "transient group" or "transient stroke model"). Baseline or "health" measurements are obtained using a portable MR system. Once the baseline measurement is completed, the brain is accessed through craniotomy. Electrocautery was used to locate and occlude Middle Cerebral Artery (MCA) for permanent groups (permanent injury model) or microaneurysm clips for transient groups. Before further measurements are taken, the bone is restored and the channel is closed.

Animals were monitored using a portable MR system for a total of 4 hours. In the transient group, after 2 hours, the brain was again reached and the clip was removed, the MCA was opened, and then the craniotomy was closed and monitoring continued using the portable MR system. To ensure future accurate placement and repositioning of the portable MR system, a veterinary wrap is used to securely attach a custom 3D printing plate to the head of each animal so that the portable MR system can be easily aligned for future monitoring.

Immediately after low field MR monitoring, the animals were taken to an MRI suite (3T Siemens Skyra). Examples of MRI scans include, by way of non-limiting example: t (T) ₁ Weighted anatomy, T ₂ Weighted FLAIR (SPACE), gadobutrol contrast Dynamic Contrast Enhancement (DCE) (TWIST), magnetic Resonance Angiography (MRA), T ₁ Mapping (VFAVIBE) and Diffusion Weighted Imaging (DWI). The MRI images provide true values for comparison with the results obtained by the portable MR system used in this experiment, which measures T ₂ Apparent dispersion coefficient (ACD) and perfusion. In this experiment, the MRI procedure takes about 60 minutes to complete.

In this experiment, an MRI scan was performed on a 48-channel 3T Siemens Magnetom Skyra (Siemens Healthcare, erlangen, germany) with a rear 20-channel head coil. T is obtained using (TE/TR=1.98/5.06/ms) and flip angle (12 DEG) ₁ Weighted DCE-MRI time series. Injection is started after the second DCE scan is acquired in order to acquire baseline signals before contrast enhancement. Gadolinium-based contrast agent (Gadobutrol, trade name Gadovist, bayer, australia) is administered through an intravenous catheter (20 g, terumosuflo) placed in the jugular vein. Gadolinium (0.1 mL/kg;3 mL/sec) was administered as a bolus using a power injector, followed by flushing with saline (0.5 mL/kg;5 mL/sec).

DCE-MRI analysis is performed using several tools, including 3D slicers, imagej, rocketship, dcemri. Jl, and custom python code following practices well known in the art, including, for example, but not limited to:

·B ₁ estimating field inhomogeneity;

·T _1,0 calculating a map;

conversion from DCE signal image to R ₁ ；

Slave R ₁ Converting to contrast agent concentration;

determining an Arterial Input Function (AIF) estimate;

conversion from AIF to Cp;

pharmacokinetic models fitted to the time series data.

Results

Data obtained from a permanent set (permanent injury model) of portable MR systems show a decrease in Apparent Diffusion Coefficient (ADC), tissue perfusion (blood replacement fraction), and T ₂ This is an increase, which is consistently related to the MRI image (as shown in fig. 29A). The mirrored hemisphere was not surgically touched and was used as a healthy or baseline reference.

The transient group (transient stroke pattern) simulates a stroke, followed by a thrombotic recanalization. Data obtained from a short set of portable MR systems show that the Apparent Diffusion Coefficient (ADC) and tissue perfusion were temporarily lowered when the aneurysm clip was applied (stroke phase) (as shown in fig. 29B).

In contrast, MRI data did not show any difference at the 5 hour time point, at which time the clip was removed and the signal had recovered. This is clinically expected and relevant to previous studies of transient cerebral ischemia, such as shown in Dorsten et al (2002) and Moseley et al (1990).

Although the apparatus and methods of this disclosure have been described in terms of embodiments included herein, it will be apparent to those of skill in the art that variations may be applied to the features or in the entirety of the apparatus and/or methods described herein without departing from the concept, spirit and scope of the disclosure. All such similar substitutes and modifications apparent to those skilled in the art are deemed to be within the scope and concept of the disclosure as defined by the appended claims.

Reference to the literature

F.A.van Dorsten et al.,“Dynamic changes of ADC,perfusion,and NMR relaxation parameters in transient focal ischemia of rat brain,”Magnetic Resonance in Medicine,vol.47,no.1,pp.97–104,2002,doi:10.1002/mrm.10021.

M.E.Moseley et al.,“Early detection of regional cerebral ischemia in cats:Comparison of diffusion-and T2-weighted MRI and spectroscopy,”Magnetic Resonance in Medicine,vol.14,no.2,pp.330–346,1990,doi:10.1002/mrm.1910140218。

Claims

1. A perfusion measurement system, comprising:

2. The perfusion measurement system of claim 1, wherein signal data from the pulse sequences are compared to provide an estimate or measurement of at least one of flow, velocity, or perfusion.

3. The perfusion measurement system of claim 1 or claim 2, wherein each of the reverse recovery RF pulse sequences includes the reverse pulse, an excitation pulse, and a karl-pezier-Mei Bum-gill (CPMG) harvest.

4. The perfusion measurement system according to any one of the preceding claims, wherein the bandwidth of the reverse recovery RF pulse sequence other than the reverse pulse is substantially the same.

5. The perfusion measurement system according to any one of the preceding claims, wherein the inversion pulses for the first inversion recovery RF pulse sequence have a narrower bandwidth than the inversion pulses for the second inversion recovery RF pulse sequence.

6. The perfusion measurement system of claim 5, wherein the first reverse recovery RF pulse sequence affects an acquisition volume within the test tissue.

7. The perfusion measurement system of claim 5 or claim 6, wherein the inversion pulses of the second inversion recovery RF pulse sequence are configured to invert Nuclear Magnetic Resonance (NMR) spins in a control volume.

8. The perfusion measurement system of claim 7, wherein the control volume exceeds the acquisition volume.

9. The perfusion measurement system of claim 7 or claim 8, wherein the control volume is configured to provide inversion buffering.

10. The perfusion measurement system according to any one of the preceding claims, wherein the inversion pulse comprises or is configured as a composite pulse.

11. The perfusion measurement system according to any one of the preceding claims, wherein the system is portable.

12. The perfusion measurement system of claim 11, wherein the portable system weighs less than about 30kg.

13. The perfusion measurement system of claim 11 or claim 12, wherein the portable system weighs less than about 25kg.

14. A method of measuring perfusion, comprising:

applying a first reverse recovery RF pulse sequence and a second reverse recovery RF pulse sequence to the test tissue, wherein the reverse pulses of the first pulse sequence or the second pulse sequence have a different bandwidth than the corresponding reverse pulses of the other of the first pulse sequence or the second pulse sequence;

15. The method of claim 14, further comprising comparing signal data from the pulse sequence to provide an estimate or measurement of at least one of flow, velocity, or perfusion.

16. The method of claim 14 or claim 15, wherein each of the reverse recovery RF pulse sequences comprises the reverse pulse, an excitation pulse, and a CPMG acquisition.

17. The method of any of claims 14 to 16, further comprising configuring bandwidths of reverse recovery RF pulse sequences other than the reverse pulse to be substantially the same.

18. The method of any of claims 14 to 17, further comprising configuring the inversion pulses of the first inversion recovery RF pulse train to have a narrower bandwidth than the inversion pulses of the second inversion recovery RF pulse train.

19. The method of claim 18, further comprising configuring a bandwidth of the first reverse recovery RF pulse sequence to affect an acquisition volume within the test tissue.

20. The method of claim 18 or claim 19, further comprising configuring a bandwidth of inversion pulses of the second inversion recovery RF pulse sequence to invert NMR spins in a control volume.

21. The method of claim 20, wherein the control volume exceeds the acquisition volume.

22. The method of claim 20 or claim 21, further comprising configuring the control volume to provide inversion buffering.

23. The method of any of claims 14 to 22, further comprising configuring the inversion pulse as a composite pulse.

24. The method of claim 23, wherein the composite pulse comprises two 90 degree pulses separated by a 180 degree pulse.

25. The method of any one of claims 14 to 24, wherein the method further comprises the step of determining whether reduced perfusion occurs compared to a clinically acceptable level.

26. The method of any one of claims 13 to 24, wherein the subject is a human.

27. The perfusion measurement system according to any one of claims 1 to 13, or the method according to any one of claims 14 to 26, wherein the test tissue to be examined is the brain, breast, kidney, liver or skin of the subject.

28. A method of diagnosing reduced blood flow or perfusion below a clinically acceptable level in a subject, the method using a measurement system as defined in any one of claims 1 to 13, comprising:

Applying a non-uniform static magnetic field to the subject at a test site;

applying a first reverse recovery RF pulse sequence and a second reverse recovery RF pulse sequence at the test site, wherein the reverse pulses of the first pulse sequence or the second pulse sequence have a different bandwidth than the corresponding reverse pulses of the other of the first pulse sequence or the second pulse sequence;

acquiring magnetic resonance signal data from the pulse sequence and processing the magnetic resonance signal data to provide perfusion data;

and determining whether the perfusion data of the test site is below a clinically acceptable level.

29. The method of claim 28, wherein the subject is a human.

30. The method of claim 27 or claim 28, wherein the test site to be examined is the brain, breast, kidney, liver or skin of the subject.